parsers: lines: support "rules" for multiple sets of regexes

TUTORIAL.md

@@ -130,6 +130,10 @@ This parser allows parsing selected invoice section as a set of lines
 sharing some pattern. Those can be e.g. invoice items (good or services)
 or VAT rates.
 
+Some companies may use multiple formats for their line-based data. In
+such cases multiple sets of parsing regexes can be added to the `rules`.
+Results from multiple `rules` get merged into a single array.
+
 It replaces `lines` plugin and should be preferred over it. It allows
 reusing in multiple `fields`.
 
@@ -142,6 +146,17 @@ Example for `fields`:
         end: \s+Total
         line: (?P<description>.+)\s+(?P<discount>\d+.\d+)\s+(?P<price>\d+\d+)
 
+    fields:
+      lines:
+        parser: lines
+        rules:
+          - start: Item\s+Discount\s+Price$
+            end: \s+Total
+            line: (?P<description>.+)\s+(?P<discount>\d+.\d+)\s+(?P<price>\d+\d+)
+          - start: Item\s+Price$
+            end: \s+Total
+            line: (?P<description>.+)\s+(?P<price>\d+\d+)
+
 ### Legacy regexes
 
 For non-text fields, the name of the field is important:

src/invoice2data/extract/parsers/lines.py

@@ -12,12 +12,12 @@
 DEFAULT_OPTIONS = {"line_separator": r"\n"}
 
 
-def parse(template, field, _settings, content):
+def parse_by_rule(field, rule, content):
     """Try to extract lines from the invoice"""
 
     # First apply default options.
     settings = DEFAULT_OPTIONS.copy()
-    settings.update(_settings)
+    settings.update(rule)
 
     # Validate settings
     assert "start" in settings, "Lines start regex missing"
@@ -113,6 +113,24 @@ def parse(template, field, _settings, content):
         # All lines processed, so append whatever the final current_row was to output
         lines.append(current_row)
 
+    return lines
+
+
+def parse(template, field, settings, content):
+    if "rules" in settings:
+        # One field can have multiple sets of line-parsing rules
+        rules = settings['rules']
+    else:
+        # Original syntax stored line-parsing rules in top field YAML object
+        keys = ('start', 'end', 'line', 'first_line', 'last_line', 'skip_line')
+        rules = [{k: v for k, v in settings.items() if k in keys}]
+
+    lines = []
+    for rule in rules:
+        new_lines = parse_by_rule(field, rule, content)
+        if new_lines is not None:
+            lines += new_lines
+
     types = settings.get("types", [])
     for row in lines:
         for name in row.keys():

src/invoice2data/extract/templates/com/com.amazon.aws.yml

@@ -7,16 +7,18 @@ fields:
   invoice_number: Invoice Number:\s+(\d+)
   partner_name: (Amazon Web Services, Inc\.)
   static_partner_website: aws.amazon.com
+  lines:
+    parser: lines
+    rules:
+      - start: Detail
+        end: \* May include estimated US sales tax
+        first_line: ^    (?P<description>\w+.*)\$(?P<price_unit>\d+\.\d+)
+        line: (.*)\$(\d+\.\d+)
+        last_line: VAT \*\*
 keywords:
   - Amazon Web Services
   - $
   - Invoice
-lines:
-    start: Detail
-    end: \* May include estimated US sales tax
-    first_line: ^    (?P<description>\w+.*)\$(?P<price_unit>\d+\.\d+)
-    line: (.*)\$(\d+\.\d+)
-    last_line: VAT \*\*
 options:
   currency: USD
   date_formats: